A Framework of an Extended Reading Passage Grading System for Personalised English Learning

An extended version of reading passage grading system is presented in this paper. Weakness of the previous work has been resolved by additional feature and assisting components. It applies four linguistic features: average amount of syllable, difficulty of vocabulary, combination of clause usage and a frequency of voice and tense. POS tagging is composed into a system to help on ambiguity issue in polysemy issue, clause and voice-tense recognising. Neural network is selected to generate a passage statistical model while students´ model is exploited to represent student´s individual preference. The user interface of the system is designed for the use as assisting tool for supporting a student-centred learning in English class.